Act Report

Taken as a whole, this data wouldn't have appeared this way and this accurate without the wrangling stages. I parse the data and see if there are missing or duplicate values, or incorrect values. Then I wrangled the data, to make it clean and tidy, and it was ready for analysis. We do this in order to ensure that there is no error, because wrong or dirty data affect the results of the analysis, and give false results.

Now, We have a large number of data, here shows a comprehensive overview of the numerical data in the archive, including the mean, standard deviation, the lowest value and the highest value.

Insight one

Dogs Image prediction

Insight two

This project is related to the ability of the system to predict and identify dogs images and classify the names and types of dogs. Here, the system made a wrong prediction 643 times. While correctly predicting 4,454 times. This was found out by opening links to images that the system rated "not a dog", and we found that they were actually dogs, and this was repeated for 643 times. From here, we see that the system is accurate in predicting correctly by 87%.

Insight three

image.png

The system classified dogs according to their types, and at the bottom it shows how the system ranked the number of dogs by type, and the Golden Retriever appears as the most type of dog that appeared in the pictures, followed by the Labrador Retriever, and so on in succession, and the miniature pinscher appears in the last

image.png